The present study compared two barrier membranes, ePTFE and cellulose, used to treat Class II furcations in mandibular molars. Fifteen patients with no history of systemic diseases and presenting matched pair defects were selected. The following soft tissue measurements were taken at baseline (after the hygienic phase), and 6 months after surgery: gingival recession, probing depth, clinical attachment level, and width of keratinized tissue. At the time of membrane placement, and 6 months later (re-entry), the following hard tissue parameters were recorded: cemento-enamel junction (CEJ) to alveolar crest, CEJ to base of osseous defect, alveolar crest to base of osseous defect, and horizontal defect depth. According to the surgical protocol, the ePTFE membranes were completely covered by the flaps (subgingival placement), whereas the cellulose membranes extended 1 mm coronal to the gingival margin (supragingival placement). Healing was uneventful in all cases and membranes remained in place for 4 weeks. Data were analyzed using the Wilcoxon signed-rank test at the 5% level of significance. No statistically significant differences were found between the ePTFE and cellulose membranes, respectively (in mm): probing depth reduction (2.87±1.0 versus 3.27±1.1), gain in attachment level (2.53 ± 1.2 versus 2.8±1.3), defect fill (3.0±1.4 versus 4.0±2.3), horizontal furcation fill (2.87±1.1 versus 2.93±1.0), alveolar crest resorption (2.4±10 versus 2.73±1.2), and intrabony defect fill (0.6±1.2 versus 1.27 ±1.7). We conclude that both ePTFE and cellulose membranes are effective, not showing statistical differences in efficacy. A larger study may be necessary to show differences in efficacy and adverse effects. J Periodontol 1998;69:454–459.